spinning Globe- Groupwise problem???

[Dremzz]

At the present moment I work for a networking company and I'm doing an
install of a gigabit network. I have installed a gigabit backbone switch
with gig uplnks going to closets to a stackable switch. I then have users
off of these switches that are running groupwise 5.1. The problem we are
having is when someone tries to access groupwise it hangs and never connects
(spinning globe). I have taken sniffer traces and saw that when the user
tries to connect to groupwise the user sends a TCP window size of 0 and the
sever sends a TCP window size 3268. The user and the server never talk and
timeout. Novell hasn't admitted to anything yet we are in the middle of
getting to the problem.
What the question I'm trying to ask is has anyone upgraded there network to
a faster backbone (to gig from atm)and the user is the only bottle neck and
has anyone seen this problem with groupwise?
I have talked to another network admin and they upgraded there network to
gig and have seen a similar problems. They fixed it by moving all there
users to 10/half on the switch and it corrected it. The problem seems to be
that the groupwise server is sending so much traffic that the NIC can't
handle it anymore because with slower network, the network would buffer the
traffic. With a gigabit network there is no buffer it's all wire speed and
the NIC can't handle it. This problem is sparatic and shows up at different
times in the day. If I move there port on the switch it corrects there
problem. We all know it's not switch because this is a higher layer than
three in the OSI model.

I just want to know if anyone has seen this problem?

Any help is greatly appreciated,
Please reply to the post and e-mail , jtu...@ex-pressnet.com